Geoenergy Projects: Balancing Innovation and Risk

Geoenergy projects, like CO2 storage and geothermal energy, are increasingly mirroring the petroleum industry's approach from initial screening to operational stages. Yet, these projects aren't without their challenges. They must navigate a complex web of physical and geological constraints, all while making risk-aware decisions that impact deployment rates and climate credibility. Here’s where the real question arises: are we truly prepared to balance these scales?

The Bottlenecks Holding Back Progress

Despite their promise, geoenergy projects face significant bottlenecks that limit progress. First, there's a lack of quality data. With scarce and often biased labels, decision-making is more guesswork than science. Second, there's a tendency to overlook uncertainty, treating it as an afterthought rather than a core deliverable. Third, we see weak connections from pore to basin, including the important chemical-flow-geomechanics link. Finally, there's a glaring lack of quality assurance and governance, especially regulator-facing deployments. These aren't just technical hurdles. they're systemic issues that demand a strategic overhaul.

Machine Learning to the Rescue?

Enter machine learning (ML), which offers a glimmer of hope. Hybrid physics-ML models, probabilistic uncertainty quantification, and structure-aware representations are making waves. But let’s ask who funded the study and who stands to benefit. Are these technologies genuinely serving the broader societal good, or are they just another tool for industry insiders?

ML's potential shines in four key applications: digital twins for imaging-to-process, multiphase flow near wells, monitoring and inverse problems like deformation and microseismicity, and basin-scale portfolio management. These aren't just buzzwords. they're tangible applications that could redefine the geoenergy landscape. But the benchmark doesn't capture what matters most if the focus is solely on performance without understanding the underlying inequities.

The Path Forward

To move forward, the industry needs more than just technological solutions. We need benchmarks, validation, and reporting standards that ensure accountability and transparency. Policymakers must step up too, providing the necessary support to make ML in geoenergy not just reproducible, but defensible. It's time to look closer at who really gains from this progress.

The future of geoenergy is a story about power, not just performance. These projects can’t just be about technological advancement. they must be about equitable benefit and representation. Whose data? Whose labor? Whose benefit? Until these questions are addressed, the true potential of geoenergy remains untapped.